In my design plan, there will be around 100 clients and they will connect to the application server over internet. Application server and Database server will be on the same LAN. (lets say the pool size is 8)
So what I expect from a pooling mechanism is the managment of these 100 clients with 8 connections. But what I see on the database server is, each of my clients is creating 8 connections which will make 800 connections in the end. And you suggest that this is a design thing in glassfish.
Am I missing something here ?
[Message sent by forum member 'ardaustun']
http://forums.java.net/jive/thread.jspa?messageID=394201